A large proportion (some estimates are as high as twenty-five percent) of the electricity generated in the United States each year goes to lighting. Accordingly, there is an ongoing need to provide lighting which is more energy-efficient. It is well-known that incandescent light bulbs are very energy-inefficient light sources—about ninety percent of the electricity they consume is released as heat rather than light. Fluorescent light bulbs are more efficient than incandescent light bulbs (by a factor of about 10) but are still less efficient as compared to solid state light emitters, such as light emitting diodes.
In addition, as compared to the normal lifetimes of solid state light emitters, e.g., light emitting diodes, incandescent light bulbs have relatively short lifetimes, i.e., typically about 750-1000 hours. In comparison, light emitting diodes, for example, have typical lifetimes between 50,000 and 70,000 hours. Fluorescent bulbs have longer lifetimes (e.g., 10,000-20,000 hours) than incandescent lights, but provide less favorable color reproduction.
Color reproduction is typically measured using the Color Rendering Index (CRI Ra). CRI Ra is a modified average of the relative measurement of how the color rendition of an illumination system compares to that of a reference radiator when illuminating eight reference colors, i.e., it is a relative measure of the shift in surface color of an object when lit by a particular lamp. The CRI Ra equals 100 if the color coordinates of a set of test colors being illuminated by the illumination system are the same as the coordinates of the same test colors being irradiated by the reference radiator. Daylight has a high CRI (Ra of approximately 100), with incandescent bulbs also being relatively close (Ra greater than 95), and fluorescent lighting being less accurate (typical Ra of 70-80). Certain types of specialized lighting have very low CRI (e.g., mercury vapor or sodium lamps have Ra as low as about 40 or even lower). Sodium lights are used, e.g., to light highways—driver response time, however, significantly decreases with lower CRI Ra values (for any given brightness, legibility decreases with lower CRI Ra).
Another issue faced by conventional light fixtures is the need to periodically replace the lighting devices (e.g., light bulbs, etc.). Such issues are particularly pronounced where access is difficult (e.g., vaulted ceilings, bridges, high buildings, traffic tunnels) and/or where change-out costs are extremely high. The typical lifetime of conventional fixtures is about 20 years, corresponding to a light-producing device usage of at least about 44,000 hours (based on usage of 6 hours per day for 20 years). Light-producing device lifetime is typically much shorter, thus creating the need for periodic change-outs.
Accordingly, for these and other reasons, efforts have been ongoing to develop ways by which solid state light emitters can be used in place of incandescent lights, fluorescent lights and other light-generating devices in a wide variety of applications. In addition, where light emitting diodes (or other solid state light emitters) are already being used, efforts are ongoing to provide light emitting diodes (or other solid state light emitters) which are improved, e.g., with respect to energy efficiency, color rendering index (CRI Ra), contrast, wall plug efficiency (lm/W), and/or duration of service.
A variety of solid state light emitters are well-known. For example, one type of solid state light emitter is a light emitting diode.
Light emitting diodes are semiconductor devices that convert electrical current into light. A wide variety of light emitting diodes are used in increasingly diverse fields for an ever-expanding range of purposes.
More specifically, light emitting diodes are semiconducting devices that emit light (ultraviolet, visible, or infrared) when a potential difference is applied across a p-n junction structure. There are a number of well-known ways to make light emitting diodes and many associated structures, and the present inventive subject matter can employ any such devices. By way of example, Chapters 12-14 of Sze, Physics of Semiconductor Devices, (2d Ed. 1981) and Chapter 7 of Sze, Modern Semiconductor Device Physics (1998) describe a variety of photonic devices, including light emitting diodes.
The expression “light emitting diode” is used herein to refer to the basic semiconductor diode structure (i.e., the chip). The commonly recognized and commercially available “LED” that is sold (for example) in electronics stores typically represents a “packaged” device made up of a number of parts. These packaged devices typically include a semiconductor based light emitting diode such as (but not limited to) those described in U.S. Pat. Nos. 4,918,487; 5,631,190; and 5,912,477; various wire connections, and a package that encapsulates the light emitting diode.
As is well-known, a light emitting diode produces light by exciting electrons across the band gap between a conduction band and a valence band of a semiconductor active (light-emitting) layer. The electron transition generates light at a wavelength that depends on the band gap. Thus, the color of the light (wavelength) emitted by a light emitting diode depends on the semiconductor materials of the active layers of the light emitting diode.
Although the development of light emitting diodes has in many ways revolutionized the lighting industry, some of the characteristics of light emitting diodes have presented challenges, some of which have not yet been fully met. For example, the emission spectrum of any particular light emitting diode is typically concentrated around a single wavelength (as dictated by the light emitting diode's composition and structure), which is desirable for some applications, but not desirable for others, (e.g., for providing lighting, such an emission spectrum provides a very low CRI Ra).
Because light that is perceived as white is necessarily a blend of light of two or more colors (or wavelengths), no single light emitting diode junction has been developed that can produce white light efficiently. “White” LED lamps have been produced which have a light emitting diode pixel/cluster formed of respective red, green and blue light emitting diodes. Another “white” LED lamp which has been produced includes (1) a light emitting diode which generates blue light and (2) a luminescent material (e.g., a phosphor) that emits yellow light in response to excitation by light emitted by the light emitting diode, whereby the blue light and the yellow light, when mixed, produce light that is perceived as white light.
In addition, the blending of primary colors to produce combinations of non-primary colors is generally well understood in this and other arts. In general, the 1931 CIE Chromaticity Diagram (an international standard for primary colors established in 1931), and the 1976 CIE Chromaticity Diagram (similar to the 1931 Diagram but modified such that similar distances on the Diagram represent similar perceived differences in color) provide useful reference for defining colors as weighted sums of primary colors.
The CRI Ra of efficient white LED lamps is generally low (in the range 65-75) as compared to incandescent light sources (CRI Ra of 100). Additionally, the color temperature for LEDs is generally “cooler” (˜5500K) and less desirable than the color temperature of incandescent or CCFL bulbs (˜2700K). Both of these deficiencies in LEDs can be improved by the addition of other LEDs and/or luminescent material(s) of selected saturated colors. As indicated above, light sources according to the present inventive subject matter can utilize specific color “blending” of light sources of specific (x,y) color chromaticity coordinates (U.S. Patent Application No. 60/752,555, filed Dec. 21, 2005, entitled “LIGHTING DEVICE AND LIGHTING METHOD” (inventors: Antony Paul Van de Ven and Gerald H. Negley) and U.S. patent application Ser. No. 11/613,714, filed Dec. 20, 2006 (now U.S. Patent Publication No. 2007/0139920), the entireties of which are hereby incorporated by reference as if set forth in their entireties). For example, light from additional selected saturated sources can be mixed with the unsaturated broad spectrum source(s) to provide uniform illumination without any areas of discoloration; and if desired, for cosmetic reasons, the individual light emitters can be made to be not visible as discreet devices or discreet color areas when the illumination source or aperture is viewed directly.
Light emitting diodes can thus be used individually or in any combinations, optionally together with one or more luminescent material (e.g., phosphors or scintillators) and/or filters, to generate light of any desired perceived color (including white). Accordingly, the areas in which efforts are being made to replace existing light sources with light emitting diode light sources, e.g., to improve energy efficiency, color rendering index (CRI Ra), wall plug efficiency (lm/W), and/or duration of service, are not limited to any particular color or color blends of light.
Aspects related to the present inventive subject matter can be represented on either the 1931 CIE (Commission International de I'Eclairage) Chromaticity Diagram or the 1976 CIE Chromaticity Diagram. Persons of skill in the art are familiar with these diagrams, and these diagrams are readily available (e.g., by searching “CIE Chromaticity Diagram” on the internet).
The CIE Chromaticity Diagrams map out the human color perception in terms of two CIE parameters x and y (in the case of the 1931 diagram) or u′ and v′ (in the case of the 1976 diagram). For a technical description of CIE chromaticity diagrams, see, for example, “Encyclopedia of Physical Science and Technology”, vol. 7, 230-231 (Robert A Meyers ed., 1987). The spectral colors are distributed around the edge of the outlined space, which includes all of the hues perceived by the human eye. The boundary line represents maximum saturation for the spectral colors. As noted above, the 1976 CIE Chromaticity Diagram is similar to the 1931 Diagram, except that the 1976 Diagram has been modified such that similar distances on the Diagram represent similar perceived differences in color.
In the 1931 Diagram, deviation from a point on the Diagram can be expressed either in terms of the coordinates or, alternatively, in order to give an indication as to the extent of the perceived difference in color, in terms of MacAdam ellipses. For example, a locus of points defined as being ten MacAdam ellipses from a specified hue defined by a particular set of coordinates on the 1931 Diagram consists of hues which would each be perceived as differing from the specified hue to a common extent (and likewise for loci of points defined as being spaced from a particular hue by other quantities of MacAdam ellipses).
Since similar distances on the 1976 Diagram represent similar perceived differences in color, deviation from a point on the 1976 Diagram can be expressed in terms of the coordinates, u′ and v′, e.g., distance from the point=(Δu′2+Δv′2)1/2, and the hues defined by a locus of points which are each a common distance from a specified hue consist of hues which would each be perceived as differing from the specified hue to a common extent.
The chromaticity coordinates and the CIE chromaticity diagrams are explained in detail in a number of books and other publications, such as pages 98-107 of K. H. Butler, “Fluorescent Lamp Phosphors” (The Pennsylvania State University Press 1980) and pages 109-110 of G. Blasse et al., “Luminescent Materials” (Springer-Verlag 1994), both incorporated herein by reference.
The chromaticity coordinates (i.e., color points) that lie along the blackbody locus obey Planck's equation: E(λ)=A λ−5/(e(B/T)−1), where E is the emission intensity, λ is the emission wavelength, T the color temperature of the blackbody and A and B are constants. Color coordinates that lie on or near the blackbody locus yield pleasing white light to a human observer. The 1976 CIE Diagram includes temperature listings along the blackbody locus. These temperature listings show the color path of a blackbody radiator that is caused to increase to such temperatures. As a heated object becomes incandescent, it first glows reddish, then yellowish, then white, and finally blueish. This occurs because the wavelength associated with the peak radiation of the blackbody radiator becomes progressively shorter with increased temperature, consistent with the Wien Displacement Law. Illuminants which produce light which is on or near the blackbody locus can thus be described in terms of their color temperature.
Also depicted on the 1976 CIE Diagram are designations A, B, C, D and E, which refer to light produced by several standard illuminants correspondingly identified as illuminants A, B, C, D and E, respectively.
A wide variety of luminescent materials (also known as lumiphors or luminophoric media, e.g., as disclosed in U.S. Pat. No. 6,600,175, the entirety of which is hereby incorporated by reference) are well-known and available to persons of skill in the art. For example, a phosphor is a luminescent material that emits a responsive radiation (e.g., visible light) when excited by a source of exciting radiation. In many instances, the responsive radiation has a wavelength which is different from the wavelength of the exciting radiation. Other examples of luminescent materials include scintillators, day glow tapes and inks which glow in the visible spectrum upon illumination with ultraviolet light.
Luminescent materials can be categorized as being down-converting, i.e., a material which converts photons to a lower energy level (longer wavelength) or up-converting, i.e., a material which converts photons to a higher energy level (shorter wavelength).
Inclusion of luminescent materials in LED devices has been accomplished by adding the luminescent materials to a clear or substantially transparent encapsulant material (e.g., epoxy-based, silicone-based, glass-based or metal oxide-based material) as discussed above, for example by a blending or coating process.
For example, U.S. Pat. No. 6,963,166 (Yano '166) discloses that a conventional light emitting diode lamp includes a light emitting diode chip, a bullet-shaped transparent housing to cover the light emitting diode chip, leads to supply current to the light emitting diode chip, and a cup reflector for reflecting the emission of the light emitting diode chip in a uniform direction, in which the light emitting diode chip is encapsulated with a first resin portion, which is further encapsulated with a second resin portion. According to Yano '166, the first resin portion is obtained by filling the cup reflector with a resin material and curing it after the light emitting diode chip has been mounted onto the bottom of the cup reflector and then has had its cathode and anode electrodes electrically connected to the leads by way of wires. According to Yano '166, a phosphor is dispersed in the first resin portion so as to be excited with the light A that has been emitted from the light emitting diode chip, the excited phosphor produces fluorescence (“light B”) that has a longer wavelength than the light A, a portion of the light A is transmitted through the first resin portion including the phosphor, and as a result, light C, as a mixture of the light A and light B, is used as illumination.
As noted above, “white LED lights” (i.e., lights which are perceived as being white or near-white) have been investigated as potential replacements for white incandescent lamps. A representative example of a white LED lamp includes a package of a blue light emitting diode chip, made of indium gallium nitride (InGaN) or gallium nitride (GaN), coated with a phosphor such as YAG. In such an LED lamp, the blue light emitting diode chip produces an emission with a peak wavelength of about 450 nm, and the phosphor produces yellow fluorescence with a peak wavelength of about 550 nm on receiving that emission. For instance, in some designs, white light emitting diode lamps are fabricated by forming a ceramic phosphor layer on the output surface of a blue light-emitting semiconductor light emitting diode. Part of the blue ray emitted from the light emitting diode chip passes through the phosphor, while part of the blue ray emitted from the light emitting diode chip is absorbed by the phosphor, which becomes excited and emits a yellow ray. The part of the blue light emitted by the light emitting diode which is transmitted through the phosphor is mixed with the yellow light emitted by the phosphor. The viewer perceives the mixture of blue and yellow light as white light. Another type uses a blue or violet light emitting diode chip which is combined with phosphor materials that produce red or orange and green or yellowish-green light rays. In such a lamp, part of the blue or violet light emitted by the light emitting diode chip excites the phosphors, causing the phosphors to emit red or orange and yellow or green light rays. These rays, combined with the blue or violet rays, can produce the perception of white light.
As also noted above, in another type of LED lamp, a light emitting diode chip that emits an ultraviolet ray is combined with phosphor materials that produce red (R), green (G) and blue (B) light rays. In such an LED lamp, the ultraviolet ray that has been radiated from the light emitting diode chip excites the phosphor, causing the phosphor to emit red, green and blue light rays which, when mixed, are perceived by the human eye as white light. Consequently, white light can also be obtained as a mixture of these light rays.
In substituting light emitting diodes for other light sources, e.g., incandescent light bulbs, packaged LEDs have been used with conventional light fixtures, for example, fixtures which include a hollow lens and a base plate attached to the lens, the base plate having a conventional socket housing with one or more contacts which is electrically coupled to a power source. For example, LED light bulbs have been constructed which comprise an electrical circuit board, a plurality of packaged LEDs mounted to the circuit board, and a connection post attached to the circuit board and adapted to be connected to the socket housing of the light fixture, whereby the plurality of LEDs can be illuminated by the power source.
There exist “white” LED light sources which are relatively efficient but which have poor color rendering, typically having CRI Ra values of less than 75, and which are particularity deficient in the rendering of red colors and also to a significant extent deficient in green. This means that many things, including the typical human complexion, food items, labeling, painting, posters, signs, apparel, home decoration, plants, flowers, automobiles, etc. exhibit odd or wrong color as compared to being illuminated with an incandescent light or natural daylight. Typically, such white LED lamps have a color temperature of approximately 5000K, which is generally not visually comfortable for general illumination, which may, however, be desirable for the illumination of commercial produce or advertising and printed materials.
Colored objects illuminated by RGB LED lamps sometimes do not appear in their true colors. For example, an object that reflects only yellow light, and thus that appears to be yellow when illuminated with white light, may appear de-saturated and grayish when illuminated with light having an apparent yellow color, produced by the red and green LEDs of an RGB LED fixture. Such lamps, therefore, are considered not to provide excellent color rendition, particularly when illuminating various settings such as in general illumination, and particularly with regard to natural sources. In addition, currently available green LEDs are relatively inefficient, and thus limit the efficiency of such lamps.
Some so-called “warm white” LEDs have a more acceptable color temperature (typically 2700 to 3500 K) for indoor use, and in some cases, many (but not all) of such warm white LEDs have good CRI Ra (in the case of a yellow and red phosphor mix, as high as Ra=95), but their efficacy is generally significantly less than that of the standard “cool white” LEDs.
Employing LEDs having a wide variety of hues would similarly necessitate use of LEDs having a variety of efficiencies, including some with low efficiency, thereby reducing the efficiency of such systems and dramatically increasing the complexity and cost of the circuitry to control the many different types of LEDs and maintain the color balance of the light.
There is therefore a need for a high efficiency solid-state white light source that combines the efficiency and long life of white LED lamps with an acceptable color temperature and good color rendering index, good contrast, a wide gamut and simple control circuitry.
In the case of conventional LED packages which include a phosphor, a significant proportion (e.g., in many cases, as much as 20% to 25%) of the excitation light (i.e., light from the LED) is reflected (back-scattered) from the phosphor back into the light emitting diode chip/package. Back-scattered light which is scattered back into the light emitting diode chip itself has a very low probability of coming out of the chip, and hence, such back-scattering results in a system loss of energy.
In addition, the phosphor converted light is omnidirectional, so that in general, 50% of the light is directed back to the LED source.
Furthermore, if the luminescent element is too thick, and/or if the luminescent material (e.g., phosphor) content in the luminescent element is too great, “self-absorption” may occur. Self-absorption occurs when light emissions within the packaging layer stay within the packaging layer to excite other phosphor particles and eventually are absorbed or are otherwise prevented from exiting the device, thus reducing performance (intensity) and efficiency. Additionally, if the particle size of the luminescent material (e.g., phosphors) is too large, the particles of luminescent material can cause unwanted scattering of both the excitation source (the LED chip) and the light generated by the phosphor.
There is an ongoing need for ways to use light emitting diodes in a wider variety of applications, with greater energy efficiency, with improved color rendering index (CRI Ra), with improved wall plug efficiency (lm/W), lower cost, and/or with longer duration of service.